Andy-at-home: anyone here have an HD3850 X2?
marvin24_: how to find out which texture format an application uses?
ROKO__: hello
ROKO__: about radeon can tell me why freeze the screen and block machine .... gentoo only have this problem ....
soreau: Are you using kms?
ROKO__: think no
ROKO__: ^^
ROKO__: CONFIG_DRM_KMS_HELPER=m
ROKO__: # CONFIG_DRM_RADEON_KMS is not set
ROKO__: CONFIG_DRM_I915_KMS=y
ROKO__: ?
soreau: Which card do you have?
ROKO__: radeon x300
soreau: You probably want FRAMEBUFFER_CONSOLE and DRM_RADEON_KMS
ROKO__: I want to work and not blocked at work
ROKO__: what to do need
ROKO__: ?
ROKO__: CONFIG_FRAMEBUFFER_CONSOLE=y
soreau: in addition, you will want to build libdrm, mesa, xf86-video-ati and xorg-server with VIDEO_CARDS=radeon in make.conf
ROKO__: Installed versions: 2.4.20!t(13,45,25 11.05.2010)(kernel_linux -debug)
ROKO__: its libdrm
ROKO__: [ebuild R ] media-libs/mesa-7.8.1 USE="nptl xcb -debug (-gallium) -motif -pic (-selinux)" VIDEO_CARDS="radeon -intel -mach64 -mga -none -nouveau -r128 -radeonhd -savage -sis -svga -tdfx -via" 5,018 kB
ROKO__: mesa is allready compillied with radeon flag
soreau: If you already have DRM_RADEON, then you should be able to boot with radeon.modeset=1 as a kernel parameter to enable kms
ROKO__: soreau
ROKO__: from lilo.conf
soreau: given you have a recent kernel version
ROKO__: in line append ?
soreau: I didnt know anyone still used lilo :P
soreau: What kernel version is this?
ROKO__: append="radeon.modeset=1"
ROKO__: 2.6.33
soreau: ok, should be fine then
ROKO__: okay now try it
ROKO__: root@Gentoo:/# lilo
ROKO__: Added Slackware
ROKO__: Added Gentoo *
ROKO__: nice
ROKO__: soreau
ROKO__: http://paste.pocoo.org/show/214235/
ROKO__: look this
ROKO__: soreau all okay ?
soreau: sure
soreau: did it fix the black screen?
soreau: not sure why nvidia module is loaded though
agd5f: MostAwesomeDude: did you figure out what you were asking about COL_CN?
papo: hello
ROKO__: soreau now use nvidia but want change it with old radeon
ROKO__: radeon works perfect
ROKO__: nvidia poor
papo: the feature matrix mentions that powerplay support is "mostly" done. I am wondering what is missing, whether this is supposed to work similar to the fglrx driver and it would be interesting to look at the git commits on powerplay, but I can't find any
soreau: ROKO__: You probably should blacklist the nvidia kernel module in that case
ROKO__: okay
ROKO__: directly remove from system
Droste: papo: http://git.kernel.org/?p=linux%2Fkernel%2Fgit%2Fairlied%2Fdrm-2.6.git&a=search&h=refs%2Fheads%2Fdrm-radeon-testing&st=commit&s=pm|powermanagement&sr=1
Droste: papo: http://lists.freedesktop.org/archives/dri-devel/2010-May/000492.html
papo: Droste: Awesome
papo: Droste: I guess that what keeps me from switching from fglrx is that power consumption is incredibly high... sounds like this patch may fix that issue
Droste: papo: Yes, but I don't know if the patches are in 2.6.33/34. Maybe you need to use drm-radeon-testing from http://git.kernel.org/?p=linux/kernel/git/airlied/drm-2.6.git;a=summary
papo: Droste: Ok, I'll give that a try tonight
papo: thank you very much
gormux: hi there
ROKO__: soreau:
ROKO__: FATAL: Error inserting radeon (/lib/modules/2.6.33-gentoo-r2-ExtremeBG/kernel/drivers/gpu/drm/radeon/radeon.ko): Unknown symbol in module, or unknown parameter
ROKO__: (see dmesg)
ROKO__: radeon: Unknown parameter `modset'
ROKO__: this is message from dmesg
jcristau: try spelling it correctly
jcristau: it's modeset
ROKO__: FATAL: Error inserting radeon (/lib/modules/2.6.33-gentoo-r2-ExtremeBG/kernel/drivers/gpu/drm/radeon/radeon.ko): Unknown symbol in module, or unknown parameter
ROKO__: radeon: Unknown parameter `modset'
ROKO__: ?
ROKO__: what parameter is correct
jcristau: i just told you
soreau: typo.
soreau: it is modeset, not modset
soreau: see modinfo radeon
ROKO__: I did everything as you said, but now the problem is that many fps dropped, probably because of the KMS
soreau: How many fps were you getting with the black screen?
ROKO__: soreau
ROKO__: but now works, FPS dropped significantly
ROKO__: 165 frames in 5.0 seconds = 32.982 FPS
ROKO__: 166 frames in 5.0 seconds = 33.026 FPS
ROKO__: 164 frames in 5.0 seconds = 32.756 FPS
ROKO__: previously worked at 3000 fps but blocking
soreau: well, glxgears is not a benchmark but that does seem abnormally low
soreau: What is the output of glxinfo|grep renderer?
ROKO__: direct rendering: Yes
ROKO__: but
ROKO__: Gentoo roko__ # glxinfo | grep "renderer string"
ROKO__: OpenGL renderer string: Software Rasterizer
ROKO__: ...
ROKO__: here must be MESA
ROKO__: think
ROKO__: OpenGL renderer string: Software Rasterizer
ROKO__: then X was unable to enable the radeon OpenGL driver and is using the software fallback. If this happens, see /var/log/Xorg.0.log for errors(which are prefixed with (EE)).
ROKO__: this is from gentoo wiki
ROKO__: (EE) RADEON(0): [dri] RADEONDRIGetVersion failed to open the DRM
ROKO__: [dri] Disabling DRI.
ROKO__: soreau
ROKO__: what to do
ROKO__: ?
soreau: yea, that is why I said
soreau:
ROKO__: xorg-server 1.7.7
ROKO__: ^^
ROKO__: no problem with it?
ROKO__: soreau after rebuild
ROKO__: reboot or only reset xorg
Droste: ROKO_: reset xorg
ROKO__: okay
sedzimir: what's the suggested way of debugging radeon crashes which make even SysRq fail?
ROKO__: hmm
ROKO__: soreau
ROKO__: after rebuild
ROKO__: still no mesa randering
ROKO__: Gentoo roko__ # glxinfo | grep "renderer string"
ROKO__: OpenGL renderer string: Software Rasterizer
soreau: pastebin your X log file to psatebin.org
ROKO__: www.extremebg.co.cc/files/Xorg.0.log
ROKO__: so ?
soreau: Looks like you dont have drm>radeon built for your kernel
soreau: Look at dmesg|grep drm
soreau: could be that or libdrm not properly installed
soreau: also looks like your ddx wasnt built against libdrm with kms support enabled
ROKO__: Gentoo roko__ # dmesg | grep drm
ROKO__: [drm] Initialized drm 1.1.0 20060810
ROKO__: [drm] radeon kernel modesetting enabled.
soreau: then you need to build libdrm with kms support I suspect
ROKO__: [ebuild R ] x11-libs/libdrm-2.4.20 USE="-debug" 0 kB
soreau: huh
ROKO__: not kms flag
soreau: well maybe BioTube will help you www.extremebg.co.cc/files/Xorg.0.log
soreau: or adamk_
ROKO__: :<
ROKO__: adamk
ROKO__: BioTube
soreau: (II) [KMS] drm report modesetting isn't supported.
ROKO__: and ?
soreau: this indicates your ddx (xf86-video-ati) wasnt built with kms support
ROKO__: It will work
Droste: ROKO_: did you boot with radeon.modeset=1?
soreau: When you build this package, it should say Kernel modesetting: Yes
ROKO__: [ebuild R ] x11-drivers/xf86-video-ati-6.13.0 0 kB
ROKO__: no use flags for xf86-video-ati
ROKO__: Droste yes
soreau: just because it is installed doesnt mean it is built correctly
Droste: ROKO_: could you pastebin the full output of dmesg | grep -E "drm|radeon"?
soreau: also just looking at flags wont help anything
ROKO__: http://paste.pocoo.org/show/214280/
soreau: You need to rebuild in this order: libdrm, xf86-video-ati, mesa
ROKO__: soreau already precompiled them 4 times ....
Droste: ok, build your kernel without radeonfb and retry
soreau: ah radeonfb is screwing you
soreau: You dont want radeonfb
ROKO__: why do not you tell me before?
soreau: Because I never saw anything that indicated this
soreau: until now :)
soreau: Droste: Cant he just blacklist the module?
soreau: or rm it :P
ROKO__: soreau how blacklist it ?
soreau: I guess you need to update your kernel config too though
ROKO__: I do not compile kernel again
soreau: probably best to listen to Droste ;)
Droste: hm never tried it :-) it's safer to just not build it.
ROKO__: means to recompile?
Droste: if it's a module you can just delete it, if not recompile, yes
ROKO__: Droste how
Droste: how what? recompile?
ROKO__: how remove module
ROKO__: Gentoo roko__ # locate radeonfb
ROKO__: ops
ROKO__: /lib64/modules/2.6.33-gentoo-r2-ExtremeBG/kernel/drivers/video/aty/radeonfb.ko
Droste: yes delete the file
ROKO__: and update-modules
ROKO__: ?
ROKO__: aty128fb.ko atyfb.ko radeonfb.ko
ROKO__: all modules
ROKO__: or only radeonfb
Droste: delete all
Droste: and reboot
sedzimir: ROKO__: update-modules is rather no longer necessary
ROKO__: okay rebooting
soreau: sedzimir: I think he is stuck using some older things, like... lilo for example :P
Droste: soreau: if the output of dmesg | grep drm is only 2-5 lines then something went wrong at boot time and most of the time because another fb driver was loaded before :-)
sedzimir: soreau: I use lilo too
soreau: Droste: yea but no telling if that was the complete output or not
sedzimir: and that has nothing to do with modules
ROKO__: Gentoo roko__ # glxinfo | grep "renderer string"
ROKO__: OpenGL renderer string: Mesa DRI R300 (RV380 5B60) 20090101 TCL DRI2
ROKO__: ^^
soreau: yay
soreau: Droste: I was doing a real half a**ed job on this here ;)
Droste: ;-)
ROKO__: but still slow
ROKO__: 88 frames in 5.0 seconds = 37.514 FPS
ROKO__: 188 frames in 5.0 seconds = 37.514 FPS
ROKO__: ...
ROKO__: http://paste.pocoo.org/show/214283/
ROKO__: here is my xorg.conf
Droste: glxgears is not a benchmark ;-) kwin compositing? compiz?
ROKO__: before I showed 3000 fps and now cut to glxgears
soreau: ROKO__: you shouldnt even need an xorg.conf
ROKO__: I need something to edit xorg.conf
ROKO__: ?
Droste: yeah, but the config is ok :-) nothing exciting in there
soreau: wonders if setting LIBGL_ALWAYS_INDIRECT would make any difference
ROKO__: how ? soreau
Droste: ROKO_: kwin compositing? compiz? or any other desktop effects?
ROKO__: Droste wait
sedzimir: what's TCL there?
Droste: texture clipping and lighting
sedzimir: OpenGL renderer string: Mesa DRI R300 (RS400 5A62) 20090101 NO-TCL
sedzimir: should I worry?
ROKO__: quake3 will try to know before that worked perfectly with "radeon"
Droste: sedzimir: only if rs400 has tcl support :-) but since this is an IGP it's possible that it doesn't.
ROKO__: not a lot of coin, it seems still not right
Droste: pastebin Xorg.log and dmesg | grep -E "drm|radeon"
ROKO__: www.extremebg.co.cc/files/Xorg.0.log
ROKO__: http://paste.pocoo.org/show/214285/
Droste: no, everything is right :-)
ROKO__: hm
Droste: some games are still slower with dri2 and kms. quake3 could be one of them
ROKO__: but why so fell 3D acceleration?
Droste: (you can always switch back with radeon.modeset=0)
ROKO__: went before
MostAwesomeDude: agd5f: No, but it doesn't look like I need it right now. I can poke at it later.
ROKO__: Even now under Slackware with the same driver
Droste: ROKO_: you could try to add Option "ColorTiling" "true" under section "Device" to your xorg.conf. but I don't know if it helps with quake3
Droste: or boot with radeon.modeset=0 to go back to UMS
ROKO__: okay
ROKO__: Option "AccelMethod" "EXA"
ROKO__: Droste soreau
ROKO__: http://extremebg.co.cc/files/crash.jpg
ROKO__: http://extremebg.co.cc/files/crash2.jpg
ROKO__: talking about it, worked for some time and block and do not know why, just have this problem in Gentoo
Droste: you could try a more recent kernel
ROKO__: I do not think this is a problem on slackware with 2.6.29
ROKO__: and works without problem
Droste: ROKO_: it is not necessary a problem with the graphics card if the whole system locks up. maybe a regression in 2.6.33 which is fixed in 2.6.43-rc7
Droste: but I'm just guessing
ROKO__: Droste
ROKO__: I did not know that 2.6.43 is released
Droste: 2.6.34
Droste: sorry
ROKO__: you drive me to compile a new one?
Droste: no, it's just an option ;-) you don't have any logs after the system crashed?
Droste: Xorg.log.old?
ROKO__: wait
ROKO__: www.extremebg.co.cc/files/Xorg.0.log.old
ROKO__: going out for a while, if you find something please write to prv
Droste: wait a sec
Droste: this is a log with nvidia driver
lucx: agd5f ping
adamk: I was pinged?
lucx: hi
lucx: agd5f told on phoronix forums that exists documentation to make vdpau acc on oss drivers
lucx: can you point me to documentation?
adamk: I believe that all the documentation AMD has released is here: http://www.x.org/docs/AMD/
lucx: thank yopu
MostAwesomeDude: VDPAU could be implemented on top of the video layer of Gallium. Talk to ymanton on #dri-devel or send an email to mesa-dev.
marvin24_: is there a way to enable s3tc for r600 (I have libtcx_dxtn)?
marvin24_: maybe some SETfield lines in r600_texture.c
adamk: It just is not supported at this time.
marvin24_: yeah - I read that in the code, I though it was just disabled, because of legal issues
adamk: No, I'm pretty sure it just doesn't work :-)
marvin24_: also the drm needs to be ajusted - but no docs available :-(
cafaro: Trying to get X working with Debian Squeeze, my card is a Mobility HD3650, so far X isn't even starting..., here's the Xorg.0.log, from when I did startx: http://pastebin.com/me16LX4W
cafaro: oh and about the Radeon: acceleration initialization failed error, I got the same error with lenny, and actually worked for me
jcristau: cafaro: did you install the firmware-linux package?
cafaro: No I didn't, will try, but I don't understand why it's needed for the radeond driver..?
jcristau: the driver has to upload some microcode to the chip in order to enable acceleration
cafaro: ok
NightNord: Greetings, everyone. I have a strange question, more related to mesa itself, probably
NightNord: For a reason, i'm developing opengl application, that need to be run via xhost
cafaro: jcristau: I did actually get a lighter screen with the firmwarepackage, but still not working, let me check xorg.log...
NightNord: In my work environment, I've got a machine with very old mesa-7.0 and friends and newer machine with mesa-7.9. In this environment, network commands passing works just fine
NightNord: But it lacks of several very usefull features like VBOs, that are supported on rendering site, but not detected via networked glxinfo
cafaro: jcristau: The radeon error from xorg.log is actually gone, only the GlidePoint errors remain (see previous log), but I doubt that error would effect X' display
jcristau: cafaro: yeah those are normal
jcristau: cafaro: what kind of monitor(s) is connected?
cafaro: jcristau: Just my laptop screen
NightNord: I had an idea, that this could be because of very old mesa installed on x-client. So, I've tried to make similar forwarding with other two machines, both running mesa-7.9-devel, with radeon driver.
NightNord: But now I'm getting "bad request" x-error somehow related to dri2
jcristau: cafaro: ok, that's what the driver seems to detect..
jcristau: using a 1366x768 mode
NightNord: So, that's my question: maybe in newer dri2 architecture such mechanics are just unsupported? Or there is something that i've missed?
jcristau: cafaro: can you pastebin your xorg.conf too?
NightNord: (it fails on DRI2Connect)
cafaro: jcristau: http://pastebin.com/xerddt3N
jcristau: cafaro: same happens if you move it away?
cafaro: jcristau: Remove xorg.conf?
jcristau: yeah
jcristau: shouldn't make a difference, but..
cafaro: jcristau: same black screen
Droste: cafaro: you should use a kernel >=2.6.33
jcristau: cafaro: ok. then next thing to try would be kernel and xserver-xorg-video-ati from unstable.
cafaro: ok
Droste: NightNord: maybe #dri or #dri-devel is a better place to ask
NightNord: Ok, thanks
BeteNoire: hi, what's the equivalent channel supporting intel driver problems?
MostAwesomeDude: #intel-gfx
BeteNoire: MostAwesomeDude: thx
NTU: Hey guys i don't know why but on my custom distro, it takes about... two minutes to start up on _FIRST_ boot with KMS with radeon r700_rlc.bin and r600_rlc.bin downloaded. It says "Requesting radeon firmware" or something like that (I didn't write it down sorry) and it just hangs. Eventually KMS goes into effect and gives me a 1920x1200 console but like.. why does it do this only on my custom distro the first time I install it?
NTU: hmm.. dmesg | grep firmware says: "[drm:rv770_startup] *ERROR* Failed to load firmware!" and "disabling GPU acceleration" I'm on Kernel 2.6.33.4 (directly from kernel.org) with Prevent firmware from being built, off, and only select drivers that don't need compile-time external firmware off, and include firmwar blobs in kernel binary on. I left out the /lib/firmware directory and didn't type in radeon/R600_rlc.bin and radeon/R700_rlc.bin since I figured
mjg59: You'r failing to load the firmware, and the kernel firmware loader waits 60 seconds before giving up
Droste: NTU: is radeon a module or built in?
mjg59: The firmware images in question aren't included with the kernel, so the kernel buildsystem won't build them in without extra effort
NTU: its compiled in
Droste: then the firmware must be built in too (as far as I know)
NTU: so recompile kernel with rX00_rlc.bin with /lib/firmware directory set?
Droste: and CONFIG_FIRMWARE_IN_KERNEL=y
chithead: the exact values are listed in the wiki article in the channel topic
Droste: http://wiki.x.org/wiki/radeonBuildHowTo#TroubleshootingExtraFirmwareforR600.2BAC8-R700.2BAC8-Evergreen
NTU: cheers mates!
NTU: that worked! thanks
x3nU: hello, i'm happy with propertiary nvidia drivers on windows :*
x3nU: bye
twnqx: wtf
twnqx: well.
NTU: just a troll
egns: lulz
egns: hahahhaa
egns: good one.
NTU: what gets me is why doesn't #ati get trolled? why is it just #radeon?
twnqx: just this moment fails to use somehting he used for years... binary nvidia drivers
egns: K-Owned
NTU: heh
evil_core: I was also happy with nvidia driver, except nahgs(f*** AGP bridge)
twnqx: aaand nvidia driver problem solved through manual override in the config \o/
DanaG: It's a bummer there's not good support for multi-gpu in xrandr.
DanaG: Two heads per GPU, two GPUs.
cTn: hi guys, i am compiling new kernel from drm-radeon-testing git and i need a little help, anyone willing to give me a hnd ?
cTn: *hand
BioTube: cTn: how do you mean?
cTn: BioTube: well, i never compiled a kernel before, so iam mostly following a guide that i found
evil_core: cTn: no, we dont like strangers!
cTn: evil_core: :( ^^
soreau: hi cTn
cTn: soreau: oh, hi there :)
soreau: cTn: So you have downloaded d-r-t?
cTn: soreau: yeah (i think so), BioTube is currently helping me via PM, because i tried this alone on my laptop, kernel was building for like an hour or so, and then compilation failed, so i wanna do it right this time
soreau: ah ok, then listen to BioTube
cTn: soreau: thanx for the interest, make -j2 is running now so, we will see
evil_core: make -j3
evil_core: how to enable FSAA and multisampling generally?
evil_core: I heard MAD is doing it, but its disabled somehow
NTU: make -j5 over here!
cTn: anyone care to tell me whats the difference ? :)
BioTube: how many simultaneous compilations are going on
cTn: oooh, nice
NTU: its actually a tad bit slower than -j4 even though I have 6MB of L3 cache + a unlocked core. cTn: -jX (X = %d) = how many cores in use
cTn: well i have only c2d, so 2 -j2 is ok i think
NTU: i used -j3 on my C2D back in the day when I had one. AMD Phenom II X3 720 Black Edition + 4th core 1337 h4x0r3d to be running, pwns it across the board :P
cTn: oooh i soooo cant wait to see some error :D
evil_core: +1
NTU: i actually just compiled gcc, glibc, and all that with make -j5 test, and it was fine. btw its overclocked from 2.8 to 3.2 Ghz!
evil_core: untiul you use BrainFuckScheduler
hnsr: using -j5 on my Q6600
hnsr: kernel builds superfast with that :p
evil_core: lol
evil_core: in developers machine we got 8 cores and 64GB RAM
hnsr: oof :(
evil_core: theres everything build superfast :P
cTn: sounds good :)
evil_core: but kernel and mesa I am building locally on T60p :(
evil_core: internet cdonnection is to slow to copy it there and over back
cTn: superfast compiling still going on
cTn: :D
hnsr: same :D
hnsr: i will get some coffee though np
NTU: im just doing some intensive gcc tests like the ppl test, glibc test, and some others. :D
NTU: i want to make sure all is OK before building all my stuff
hnsr: man im seriously annoyed with my new phone
hnsr: here i was thinking all android are open and you can do anything with them you want
hnsr: and yet i have to jump through some major hoops to get root on it
hnsr: i guess i should blame the manufacturer (HTC) but still
NTU: panning > android
NTU: :D
cTn: hnsr: yeah, i own a htc hero (its my first and last android phone from htc)
hnsr: i read some guides on how to gain root and it is just scary
hnsr: have to do all kinds of weird stuff and non of it is explained anywhere
hnsr: spent the entire googling for stuff and i still don't really know what is involved :|
hnsr: s/entire/entire day/
cTn: i was one of the lucky ones that had old firmware so i used exploit to gain root, but if i am correct every moethod including gold card is using an exploit
hnsr: yeah something about a gold card here as well
hnsr: htc desire btw
cTn: did you rooted him or not yet?
hnsr: evne what a goldcard is is not explained enywhere i could find :( just that it is some 'special' sdcard
hnsr: nah not yet
cTn: *it
hnsr: im not going to root it until i actually know what im doing :p
cTn: let me take a look maybe i can help you
cTn: ow yeah i am good
cTn: hnsr: http://android.modaco.com/content/htc-desire-desire-modaco-com/307365/04-may-r4-htc-desire-rooting-guide-now-with-hboot-0-80-and-os-to-1-21-support/
cTn: have fun
hnsr: yeah ive read that one
cTn: didnt work ?
hnsr: the problem is nothing is really explained :p just a step-by-step instructions
cTn: (let me take a look again)
hnsr: like what is a rom (and a rom kitchen? :/)
cTn: oh i was thinking that you are concerned that those script files do, but its easy to find out with just editing
cTn: so
cTn: the "software"/firmware you have on your phone
cTn: is stock rom from htc
cTn: vanilla android + htc crap on top
cTn: when you gain root access you are probabbly going to flash an custom recovery image
hnsr: hmm
cTn: which will allow you to flash custom "firmware", but understand
cTn: firmware from htc = radio ( code that controls your sim card etc etc, if u mess it up, your phone is dead), system, apps, boot
cTn: oh and ofc recovery
cTn: custom roms are mostly just system + apps
cTn: so u cant really "hurt" your phone so do speak
cTn: (and my make -j2 got stuck 10 minutes ago, coool lol)
cTn: hnsr: anything specific that you are interest in knowning? :)
cTn: *interested
hnsr: hmm sec let me just re-read that :p
cTn: yeah
hnsr: ok so i should probably not touch firmware, but i can still flash custom android on it, like say 2.2 when it comes out?
hnsr: in case HTC doesn't release it in a timely manner
hnsr: and run custom binaries (like maybe a bash shell, or at least something nicer than the default 'sh')
cTn: well, the key point in this is, if htc releases an update and u flash it
cTn: the "normal" no root way
cTn: you might not gain root ever again
hnsr: ugh, i guess i better turn auto-update off :p
cTn: this root guides always use some hole in kernel to give u root access
cTn: yeah
cTn: but just so u know, if u root, and lets say your microphone broke
cTn: u can always flash "original" firmware over the custom one and send it back for repairs
hnsr: oh good, that is good to know
cTn: i jsut dont recommend touching anything "radio" related or "SPL" related unless you know what you are doing
hnsr: oh yeah, what is SPL?
hnsr: brb btw getting some coffee :p
cTn: if i understand correctly its a second program loader, android phone have multiple "boot states", normal boot, boot to recovery, fast boot, hboot or i dont know what else
cTn: and spl is a part of fastboot or hboot which checks for the "rom"/firmware signatures, but if you flash frong spl or flashing fails = dead phone, same as with radio
cTn: radio is the "first" code which phone loads, and if its broken you cant even charge your phone, and htc service will have to replace your phones mobo
NTU: brb
hnsr: cTn, ah I see
hnsr: thanks a lot for the info :p
cTn: you are welcome, i v read so many "noobie" threads with ppl that flashed radio from CDMA phone to GSM, or SPL from one phone to another, or they pulled battery out during flashing
cTn: first 5 or 6 times i was just sad for the ppl, but now (when you dont know what you are doing, dont do it!)
hnsr: yep
cTn: but i think now you know everyting important to get you going
hnsr: i do feel a bit more confident now :p
cTn: getting root is nothign to worry about, but root privileges is the thing that just some ppl dont get it, with this right u can kill your phone
cTn: oh and so u know, about the "root" using applications
cTn: when you root your phone, and probabbly flash some custom rom on it
cTn: there will be a superuser permissions manager
cTn: it works like a "firewall" you let or deny app asking for root
hnsr: ah right
cTn: so no noobie hacker which send you some infected app can hack your phone
hnsr: yeah i guess by default apps still get their own unpriveleged user
cTn: and i think this is safer with custom rom then with stock, because if the exploit in kernel still exists
cTn: any app on the market can "hack you"
hnsr: hmm
hnsr: i did see they only yesterday released the kernel sources for my phone
cTn: and the kitchen, for the kitchen use this http://forum.xda-developers.com/showthread.php?t=633246
cTn: it let you create your own custom rom (i prefer this way, because i dont trust some rom repackers, or some random guys, they could include some keyloger or whatever) you know what i mean
cTn: with this you can create everything your self and its 100% safe
hnsr: yeah
hnsr: how hard is it to roll your own rom?
cTn: with that kitchen "package" that i send you, its practically 5 min of work + flashing
cTn: everything is explianed there, you "steal" rom.zip from the original htc update package
cTn: u extract the boot and system.img files
cTn: and kitchen will take care of the rest
cTn: when you finish "cooking" this kitchen will provide you with an signed update.zip which you can just put on your sdcard and flash via your custom recovery
cTn: you can add root privilegs app via that kitchen, wifi tethering, ap2sd
cTn: etc etc etc
DanaG: argh, both #dri and #xorg are silent.
hnsr: cTn, oh wow okay, that explains a lot
DanaG: on Savage: (EE) AIGLX error: Calling driver entry point failed (EE) AIGLX: reverting to software rendering
hnsr: i didnt quite understand what the deal with the zipfiles was
hnsr: well stop the crazy offtopic talk now to make room for DanaG :p
cTn: hnsr: when you have custom recovery, it allows you to flash custom "zip" files, but they still have to be signed, call it a "hash-check" or smth like that, its just a signature, to prevent damaging your hardware by faulty zip file
cTn: ok will shut up now :)
DanaG: Anyway, my issue is off-topic for #radeon, specifically, anyway.
DanaG: But it IS on-topic for #dri and/or #xorg.
DanaG: Though, if somebody here would be able to help, go ahead. =รพ
DanaG: Not that I expect much from a Savage.
cTn: gn ppl
evil_core: DanaG: do youhave d-r-t, ddx master or mesa master?
evil_core: DanaG: its pretty normal even for radeon, and it doesnt affect performance when in noncomposited env
evil_core: DanaG: its related to hardcoded paths in AIGLX (doesnt honour LIBGL_DRIVER_PATH)
DanaG: Actually, that issue is with a Savage card.
DanaG: My radeon works wonderfully; only the Savage is failing to give any DRI at all.
evil_core: but its not related to that communicate you posted
evil_core: check, and you gor identical with radeon :P
evil_core: LIBGL_DEBUG=verbose glxgears
evil_core: I am going sleep anyway, so good night all!
NTU: :-D i wrote my first linux kernel patch! woo! now i wonder if that will effect performance much..